Wikipedia Bans AI Articles but Allows Certain Uses

Wikipedia recently updated its guidelines to prohibit the use of artificial intelligence (AI) in writing and rewriting articles. This decision, made last week, stems from the tendency of AI-generated articles to violate several fundamental content policies of the online encyclopedia.

This ban specifically applies to the English version of Wikipedia. However, editors will still be able to use AI in certain specific cases. For example, they may use language models to suggest basic corrections, provided that this does not introduce original content. AI may also be used to translate articles from other languages into English. However, editors must adhere to the site's rules regarding LLM-assisted translations, which require them to have sufficient knowledge of the original language to confirm the accuracy of the translation.

Wikipedia's new policy emphasizes that some editors may have a writing style similar to that of AI language models (LLMs). Therefore, editors will need to rely on criteria beyond mere stylistic or linguistic signs to justify restrictions on their editing capabilities. The guidelines recommend focusing on the text's compliance with content policies and the recent changes made by the concerned editor.

Wikipedia editors have been grappling with AI-generated articles for months, leading to the implementation of a "quick deletion" policy for poorly written articles. The WikiProject AI Cleanup initiative has also been formed to combat AI-written content and assist others in identifying it.

This update to the guidelines was proposed by an editor known by the pseudonym Chaotic Enby. After lengthy discussions within the community, the proposal was adopted with overwhelming support, targeting the obvious issues related to the use of LLMs while allowing for certain usages deemed acceptable.